Political reservations for women create a ‘potential of presence’ (Agarwal 2010), and offer prospects of diversity and other governance benefits. However, in settings where gender discrimination may affect the characteristics and attitudes of candidates for political office, women representatives are less likely to be politically or administratively experienced. Thus, even if women political leaders were inherently more development-oriented than their male counterparts, existing disparities might initially blur and significantly delay the onset of governance gains.
While evidence on the quality of public service delivery in gram panchayats (GPs) headed by women in India is unclear (Chattopadhyay and Duflo 2004, Ban and Rao 2008), research on the impact of female leadership on corruption in public programmes is practically absent. This fundamental neglect is of immediate policy relevance, given the renewed global commitment to increasing women’s participation in political life (World Development Report 2012) and the intense debate within India about whether to reserve state and national legislature seats for women or not.
Political reservation for women and corruption
In a recent study I, along with my co-authors, assessed the impact of women political leaders on corruption and other aspects of the quality of delivery of India’s most ambitious public programme to-date—the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) (Afridi Iversen Sharan 2012). By exploiting data both across GPs and over time in the same GPs we are able to study whether and how governance outcomes are affected as female leaders accumulate knowledge and experience.
We use data from primary survey of households, GPs and mandal level (sub-district) development officers as well as audit reports from 2006-2010 for 300 GPs on the implementation of projects under MGNREGA in Andhra Pradesh. Utilising the nationwide policy of randomly reserving one-third of village council headships for women, we find that the likelihood of inefficiencies and leakages in these projects are higher in village councils reserved for women sarpanchs. These conclusions hold both for processes for which the GP is primarily responsible as well as those for which it shares responsibilities with other programme functionaries. For instance, in female-reserved GPs, the probability of a household being asked for a bribe to obtain a job card was more than 6 percentage points higher than in unreserved GPs. Delays in wage payments due—a strong hint of administrative inefficiency—are more common in female-reserved GPs. On being asked to verify labour records, households are almost 6 percentage points more likely to find discrepancies between actual and official records of wage payments under the programme in GPs reserved for females.
Learning by doing
Looking at audit reports for the unchanged elected sarpanches and accounting for variation in the performance of GPs over time, we find that although the number of irregularities in MGNREGS implementation—in both labour and materials related component of the programme—were significantly higher in female-reserved village councils at the onset of the programme, these decline with the duration of the tenure of the head or sarpanch.
We do not find evidence supportive of reporting biases, gender stereotypes or differences in the probability of re-election of male and female sarpanches as possible explanations for these results.
However, the audit data which also contain information about functionaries on whom malfeasance in programme implementation was pinned indicate that sub-district level bureaucrats are more likely to be held accountable for discrepancies in the MGNREGS programme in female-reserved village councils. The latter suggests greater vulnerability to bureaucratic capture in village councils reserved for a woman head, a vulnerability that is likely to be aggravated by the lack of prior political and administrative experience of female sarpanches. Indeed, in instances where female sarpanches have prior political experience and are less likely to require assistance in executing their day to day duties, the governance of the programme is significantly better.
Note that the level of gender parity and women’s education in Andhra Pradesh is higher than the national average. Our estimates, therefore, are likely to represent lower bounds on the effect of women leaders on public programme implementation in India and elsewhere. Thus, in states such as Bihar and Uttar Pradesh, which have higher levels of gender discrimination, the effect of political reservations for women on the implementation of public programme may be much worse. But at the same time, the potential gains from learning by doing over time could be significantly higher.
Resonating with other studies, we also find that the probability that a programme-related grievance is made by a woman, and the number of complaints registered by women during an official audit is higher in female-reserved village councils. This suggests that the presence of a female leader strengthens women’s voices (Iyer et al. 2012).
Need for capacity building and institutional support
The findings reported here are of immediate policy relevance—political and administrative experience can not only lower corruption in public programme delivery but also expedite the onset of governance gains in constituencies headed by women. This highlights the need for capacity building and institutional support to make women’s political participation and affirmative action policies more effective. Lack of adequate administrative support and training in the utilisation of financial resources by grassroots institutions may seriously undermine the effectiveness of public programmes and of affirmative-action policies in Indian and other developing countries.
Farzana Afridi is with the Indian Statistical Institute.
“This column has been reprinted with permission from Ideas for India www.ideasforindia.in”